If you sell on Artgrid and Storyblocks your income is much higher than your income on Pond5 and other stocks too.

I would like to know what you are relying on to guarantee what you say. Just your portfolio? If it's just that, it's not enough.

Let's be modest, shall we? maybe it works for you, not for me.

The only problem is that it is almost impossible to get on Artgrid or Storyblocks,

Good thing they don't let practically anyone in. 

Quote from: cascoly on June 15, 2023, 13:02

Quote from: Evaristo tenscadisto on June 14, 2023, 15:34

...

"Generative AI, like ChatGPT, would have to comply with transparency requirements:
...
    Publishing summaries of copyrighted data used for training


... This includes AI systems that generate or manipulate image, audio or video content, for example deepfakes."

...

a good first start - I'm hope they consider:

how can those offering ai know what images were used when the systems don't provide that info & how would it be saved &H displayed?  besides working thru agencies, artists offer their work on  independent website's & POD sites such as FAA

This is not a rule for microstock sites offering AI images for sale, this is a rule for companies that develope AI. They, so sites like ChatGTP or Midjourney or stable difusion would have to show which images they used to train their AI.

or Adobe firefly which is an ai art generator too... or maybe Adobe photoshop where Adobe Unveils Future of Creative Cloud With Generative AI as a Creative Co-Pilot in Photoshop. 

Just a follow up to my last post in this topic...

The EU has come one-step closer to having the world's first law regulating Artificial Intelligence. The European Parliament voted today in favor to approve the AI Act. Next step, the talks will now begin with EU countries in the Council on the final form of the law. The aim is to reach an agreement by the end of this year.

From the series of measures, I highlight the following extracted from the news of the European Parliament:

"Generative AI, like ChatGPT, would have to comply with transparency requirements:

    Disclosing that the content was generated by AI
    Designing the model to prevent it from generating illegal content
    Publishing summaries of copyrighted data used for training

Limited risk
Limited risk AI systems should comply with minimal transparency requirements that would allow users to make informed decisions. After interacting with the applications, the user can then decide whether they want to continue using it. Users should be made aware when they are interacting with AI. This includes AI systems that generate or manipulate image, audio or video content, for example deepfakes."


link to European parliament news: https://www.europarl.europa.eu/news/en/headlines/society/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence

Quote from: wds on June 09, 2023, 13:52

I'm guessing one important factor is as follows:

  - Many photos uploaded to Stock sites aren't super saleable....i.e. snapshots, nice landscapes....etc.. Generally with stock, photos with people in "interesting" locations (e.g. offices) sell better. For a photographer, that means hiring models with releases etc., and getting those interesting locations. For AI, it is no problem to include "people" as subjects in "interesting lcations". In general, with good prompts, AI can produce saleable images without having to hire actual models or travel to interesting or challenging locations as a photographer would have to do.

Right!
The AI is a powerful tool and another resource at your disposal. Important is the choice of the subjects and the post processing. Anyway some of my AI generated images are performing very well with more than 50 downloads each and others are approaching...

Congrats! 

Quote from: Evaristo tenscadisto on June 12, 2023, 08:37

[...] AI is not replacing you "doing" images. You use AI to make your images.
[...] you should give a try and learn the new processes too. After all you use these days more the PC with email/internet and not letters with post riders.

My reason is more ethical than technical. And I can have a relevant opinion on multiple things without even having to try them.
But true, the question is doing real photographs instead of AI images, I go change the word in my post.
Whatever, you're welcome to participate to the game I proposed, since you come and write here

Game is a new form of Art. Just by raising the issue in ethical terms it paddles us towards the greatest thing about Art: the boundaries of it.
In fact, art has always been concerned with just that since Greek times. Bringing new ideas to the table that bring the world of things as we know them up for debate. But the question that i raise with your game:

Is it that important to discuss whether or not the image was created with AI assistance?
Isn't the content of the created image more important than the technical aspect that you try to get the player to identify in your game?

In fact, you try to trick the player into reflecting on the subject in a technical perspective so that he finds an ethical point. The problem is that the ethical point you want to find and discuss in your game goes against the content value of images that you put to play. In other words you ask players to identify a technical issue (AI or reality) and not really about the ideas expressed on the images themselves.
Your type of game was also done when Digital photography appeared. At the time the purpose was identifying whether the photo had been manipulated by a digital process or whether it was purely mechanical.

The result was: no ethical issue. Both were authentic images made with different processes.

Therefore your game is based on the assumption that if a player doesn't correctly identify whether an image is AI-assisted or not, it can trick the player to make him believe that the images are not authentic. But they are. They were made by humans with AI assistance tools or without them. So your game is to some extent rigged, since you only want to show the part you want - a small portion of a technical issue.

Why didn't you put images with people with 3 legs or hands with 6 fingers? Or nuns eating spaghetti or pizza which looks Bizarre? 
For instance, i find those images much more interesting and unique, almost reflecting a new age of surrealism, deep dreams or grotesque art. Also in terms of thinking the narrative of the image - something that stopped a little bit since cubism time.

Nevertheless i find your game fun. I played and the result is: no ethical issue. All are authentic/real images made by humans using different processes. AI/reality check is a feature to trick your mind.

;-)
 

Quote from: MatHayward on June 08, 2023, 09:15

Quote from: Cider Apple on June 08, 2023, 09:00

Just wondering if I can use my AI-generated images in videos and upload them to Adobe (as adobe allows images). There is a separate AI collection for images but not for video so I'm guessing not. Maybe Matt would know?

Any asset that was created with or enhanced by generative AI software must be submitted as an illustration. This is not possible with video files. We do not accept generative AI video content at this time. If that changes in the future, I'll be sure to post here in MSG.

Thanks for the question,

Mat Hayward

But I am confused now. Then why were these animations accepted and can be sold?


https://stock.adobe.com/si/video/futuristic-cosmic-animation-with-deep-space-galaxy-and-futuristic-cityscape-on-planet-bright-cosmic-animation-with-illustrations-transformations-and-metamorphose-ai-generated-cinematic-video/599697643?prev_url=detail

https://stock.adobe.com/si/video/ai-art-generation-process-timelapse-neural-network-imagining-a-digital-artwork-from-text-description-artificial-intelligence-used-in-image-creation-generative-visual-art-4k-3d-animation/528551229?prev_url=detail

https://stock.adobe.com/si/video/ai-generated-nft-crypto-art-of-impressionist-landscapes-romantic-victorian-digital-painting-image-manipulation-of-an-artificial-intelligence-cyberart-travel-into-the-mind-concept/527553471?prev_url=detail

https://stock.adobe.com/si/video/ai-generated-girl-cyborg-with-robotic-masks/593228961?prev_url=detail

Thank you!

I have no doubt that they are animations made with AI. I'm pretty sure they were made with Stable Diffusion (5.1 or 5.2). Also title immediately indicates what it is: Artificial intelligence used in image creation. Perhaps because it is such a recent process, Matt is still unaware of the possibility of uploading AI animations and just illustrations.

Evaristo tenscadisto, programs that create AI are not available to everyone. People pay a lot of money to work in them. The income from these programs may not even cover these costs.

I think you are misinformed. I cannot understand how to spend huge amounts to generate AI images. Perhaps the user who wrote can explain better because the value is much higher than practiced by AI generators. You have essentially 3 Ai generators in first league: Stable diffusion, Midjourney and Dall-e 2.

1) Stable diffusion - You can use it for free. There are a lot ways to use it:
 
              a) They got a friendly website but you can also go to NightCafe Creator which essential is the stable diffusion too
             
              b) you can install in your PC with auto1111 and use your GPU.
             
              c) Download from github and huggingface and use it with google colab to use their GPU. You got a lot youtube tutorials on how to do it.               
                 Also you can add models that people train by install them. it's free and available for everyone. although you have to pay gpu google
                 from colab after a while.

2) Midjourney. You have a website but You need Discord to generate. A lot of tutorials in youtube. it's free first images but after you pay 8$, 24$ or 48$(pro plan). 


3) Dalle 2,When you first sign up for DALL-E, you�ll receive 50 free credits that are valid for 1 month. These credits will allow you to generate up to 200 images and get a feeling for all of the features that DALL-E 2 has to offer. Once your initial trial credits expire, you�ll receive 15 new credits every month. So technically, you could continue to use DALL-E 2 forever at absolutely no cost to you.

Once you�ve used up all your free credits, you can purchase additional credits in incremental packs of 115, which cost $15.  115 credits give you 460 images, which sounds like a lot.  But remember, every single action you do in DALL-E 2 will cost you 1 credit, so those 115 credits are a lot less than you might think. If you use some of DALL-E 2�s special features such as Inpainting or Outpainting, your credits are going to evaporate into thin air. Also bear in mind that your purchased credits will expire after 12 months.


In short you can try all for free at the beginning but if you install Stable diffusion in your pc you won't pay.
 

in short, as soon as "this thing" is trained one will get 0,00000000...
same like it was with enough images video etc

Ps: may i ask what you define with good money and do you know
    how often and how many images are used and where the outlet goes?

Don't get me wrong but i think your perspective is limited to the perception of Knowledge you have about AI.

Real authentic images/video have real data/metadata which can produce augmented reality and new perception about things that humans alone cannot do or have. Generative AI Images have "fake" simulated data and are limited to just that - artistic visions about reality.

I.E. If you are training AI for safe driving i bet you won't want "fake" metadata from AI images because it will corrupt the learn of AI itself. In this case the learning curve of AI will be something called "false positive learning" or "false negative learning" depending on the measures of safe drive attributes established primarily.

However, the actual value of each Real image will be less as there is more and more about the subject portrayed in the image. As it becomes increasingly easier to produce generative AI images and their volume increases exponentially, the values of real images will be greater than those of AI images because they got real data and metadata about reality.

The good news is that the research and development industry is far superior to the Hollywood industry. To give you a clue European Commission Spends Trillions in research and only some hundred millions in programmes for Media. In this sense one of the paths for Microstock contributors will not be so much an artistic vision of reality, but rather a perspective of reality like documentary/news and that serves or has potential to be used for AI learning.

I'm not psychic but I think that in the near future AI images are here to stay but Real Images will not disappear - only part will be substituted.

In short: adapt your work/portfolio for this new reality.

 

I was very disappointed to discover that a large number (over 200) of my video collection has been duplicated and re-uploaded to Shutterstock, Pond5 and AdobeStock (and probably other stock sites but I haven't checked).
The account goes by the name CreativeVideoIdeas in various forms and there are a lot of videos on this account that are not mine, I have reached out to one other contributor who I could track down but it will no doubt have other contributors videos as well.
Anyway check that clowns account to see if your videos are on it.

Thanks for info. I checked one of the links you sent (P5). This is an exclusive account from Vietnam. When looking at the portfolio, I didn't find any of my videos but I took the time to click on some videos in the "similar" button to see what results appeared. In one case or another similar videos appeared but slightly different (with small crop or a little less time like 1 sec. or so).

i.e.
A) Creativeideas "Beautiful Rainbow, Rainbow Over The Sea, In The Waves Crashing Against The Rocks" 
Item ID: 228522480 selling 34$HD/64$4K.

B) Very similar video by other contributor, non-exclusive, Indonesia account since 2018 with video title "Rainbow Form In Waves Crashing Against Rocks On The Ocean Sea" Item ID: 204807574 selling 30$HD/50$4K.

I don't know if this helps you but I remember receiving an email and having to submit a photocopy of my ID to continue receiving my payments from P5. P5 will only let's you withdraw sales revenue after delivering ID and link to your account Payoneer/skrill. I think this measure happened during 2016 year.

Anyway my point is they can know who is Creativeidea. Also tracking CreativeIdeas P5 sales i can see this contributor had at least 53 sales. With some luck you can figure this out with P5.

All the best!

I thought it looked a bit strange. My margin was supposed to be 20% ???Then the next day I received this fees notification. I was horrified. POD sites are supposed to make their own profits from the production of the merchandise!!  I was so angry I immediately cancelled my account.  Hopefully I will get the money due to me.  I don't care - I am sick of this.

Annie i see that you might deliver your share of VAT if outside some countries. Perhaps this was an Euro zone sale?
VAT is a tax that Europeans pay to every product we purchase. Companies that make the sale (physical/online) charge the VAT directly to the customer and then they have to deliver it to the state.

Usually is 23% VAT here but it varies (0%-32%) according to type of product (basic or luxury goods) and from country to country in EU zone. If you apply my country 23% VAT to your 42 dollars (your 20% of sale) that is approximately the value that is missing.

I am only guessing here but it might be delivery of taxes for being an EU sale.

Receive an email from Redbubble today.

I only uploaded 2 images with the intention of making some gifts to offer to some friends at a party but I ended up using a cheaper local store...

Cancel my account.

Quote from: Evaristo tenscadisto on January 12, 2023, 01:37

..
In my opinion that is a bad comparison because switchboard operators don't create ideas like artists...

There are two things going on here. Art isn't going to die but jobs for commercial illustrators will (largely) as they aren�t the ones (for the most part) creating ideas. That�s the client/ art director, and they wont need the artist/ illustrator to execute those ideas any more.

Things will be different since AI causes disruption.

There is a book called "Human + machine : reimagining work in the age of AI" published in 2018 from Paul R. Daugherty, H. James Wilson. I'm not advertising it, but I personally like some of the visions or insights the book has.

AI is already or will be in every sector of business. I won't spoil the reading too but if there's repetition, replication or redundancy in the business its a clue that tasks/processes will change. One of the things that call my attention was the scalability of companies with AI and Personalization for AI in product and Service design. It means that many companies will make personalized ads to an individual or a small group of consumers instead of one for all. 

Example: instead of one "coca-cola" video  or 10 posters for publicity to all community there will be hundreds or thousands of them spread according the data of the consumer...

The increase of advertisements will be greater in quantity and will be more diverse too. This is already happening since there's UGC creators are making videos of product reviews, unbox, or ads to spread into is own community of followers.Who hasn't seen a video of a photographer reviewing a new camera on ytube yet? 

note: btw there's also a chapter with Nike as example for designer and how they worked with AI.

These were technoligical advancements that helped you in your task. AI does the art for you. It's not a tool that helps an artist, it's a replacement.

Gameover explains some points with "AI friend or foe?", on her blog with a very brief history of art.
She calls our attention to the word �artificial� with comes from the greek philosopher Aristotle in his Rhetoric: artificiality (the quality of being made or produced by human beings rather than occurring naturally.) - very nice! 

AI is a tool (1) and cannot do "the Art for you" (2).

1) It's hard in a post to explain in detail such issue regarding onto-technologies of the body but let me try to put it in this way: AI is an extension of your body not a replacement. In the same way that you don't replace your hand with a brush, you use a brush to paint. The brush doesn't paint by itself - you need a human to do it. In AI generator you need to type and work with a prompt: sort of digital code that machine can process your request.

Henry Ward Beecher once wrote, �A tool is but the extension of a man�s hand, and a machine is but a complex tool.� This fits here in line of scientific research into �embodiment�. But there's �disembodiment� in AI too.  Which basically means that AI feeds on the output to become more efficient- it collects the data of your image to improve future Artworks. So AI is a Tool with �embodiment� and �disembodiment�features.

2) AI cannot do "art" for Humans. Art is a representation or presentation of an idea in a shape/form. AI doesn't have ideas of its own, Humans do. Therefore AI cannot do Art but can be extremely efficient in drawing, playing music, etc. Best alternative is that AI it's a performer conducted by an artist. A tool to improve your Artwork just like photoshop is using with mouse and clicks interface instead of a Prompt.

A more correct comparison in advancement in technology would be for example a telephone and switchboard operators.

In my opinion that is a bad comparison because switchboard operators don't create ideas like artists. The task/job was simple connecting cables. They were a kind of cogwheel in a communication system. New technologies made it faster, better and cheaper than humans. Now we all use a Smartphone with social media.

We don't have switchboard operators anymore but how much people you think have been hired in communication companies and social media (youtube, facebook, twitter, tiktok, instagram, whatsup, signal, etc) last 20 years?

None of social media companies existed before 2004 and according to google there are 142,282 people employed in the Social Networking Sites industry in the US as of 2023. Let's not forget about UGC creators which in fact creates everyday content and get paid to feed the channel by social media, ads or product reviews...so a lot jobs here too.

I am not saying AI will replace all art, but it will certainly cause  a decrease in artists when they have a harder time to make a living from art (which was already difficult before AI art!) as AI is cheaper and microstock photography and drawings are the easiest replaced by AIs and the first to go.

Things are evolving so to expect this market not to change is to stand still in time. I do agree at some point with Gameover analysis where "plenty of �artists� will inevitably lose their jobs as soon as their customers learn how to order a piece of art directly at the source, a skilled and most likely way cheaper AI."
I think artists that see potential of AI in their work will continue the path. I already see a lot of UGC creators with thousands of followers only doing AI too.

This week i was approached twice by NFT collectors that want to buy my AI artwork so this can be a new market too. Others will go more for news/events and street life photography (there will always be in demand since AI cannot produce reality). Companies of every kind are hiring people to deal with social media, specially creatives that can do all in one: illustrations/ Shoot photos/videos and editing them.

At the end eventually people will adapt to new reality.